Buckets:
| # Run once Server A (thinking-off) is confirmed healthy. | |
| # SERVER_A_URL = e.g. https://6ac4cbe1fbc85ba6823b360c--30000.hf.jobs | |
| set -e | |
| SERVER_A_URL="$1" | |
| if [ -z "$SERVER_A_URL" ]; then | |
| echo "usage: deploy_cmd_final.sh <SERVER_A_URL>" >&2 | |
| exit 1 | |
| fi | |
| hf jobs run \ | |
| --flavor cpu-basic --expose 8080 --timeout 12h \ | |
| --label project=terminal-bench-harness --label component=think-router \ | |
| --volume hf://buckets/SaylorTwift/harness-evals:/data \ | |
| --env UPSTREAM_BASE_URL="$SERVER_A_URL" \ | |
| --env MODEL_NAME=Qwen/Qwen3.8-27B-FP8 \ | |
| --secrets "UPSTREAM_TOKEN=$(hf auth token)" \ | |
| -d python:3.12-slim \ | |
| bash -c ' | |
| pip install -q -r /data/think-router-code/requirements.txt | |
| cd /data/think-router-code | |
| uvicorn app:app --host 0.0.0.0 --port 8080 | |
| ' | |
Xet Storage Details
- Size:
- 777 Bytes
- Xet hash:
- a5a8b70f0111fd80a0bbe8e4887acf4e09154f17f4495ba4a06a3e6057db369d
·
Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.